1. Field of the Invention
Approximately 3 billion chickens are raised in the United States every year, and virtually all are contaminated with Campylobacter jejuni (C. jejuni). Currently, there is no vaccine or any other method available to the poultry industry for the prevention of colonization, therefore there exists a need to develop a product which will provide protection from C. jejuni contamination. This invention relates to a novel fusion protein which, upon administration to a poultry population, will decrease the incidence of colonization by C. jejuni. The protein is relatively simple to produce and purify, and it is expressed in large quantities and can be used without further treatment (beyond purification) for vaccination.
2. Description of the Prior Art
In recent years, C. jejuni has been recognized as a major human enteropathogen and is the species implicated in more than 95% of the cases of campylobacteriosis in the United States (Karmali et al. 1983. J. Infect. Dis., vol. 147, pp. 243-246). C. jejuni has also been recognized as a common cause of gastroenteritis worldwide (Georges-Courbot et al. 1989. Res. Microbiol. vol. 140, pp. 292-296) with two-thirds of the reported cases found in the United States (Bokkenheuser and Sutter. 1981. pp.301-310. In Diagnostic Procedures for Bacterial, Mycotic and Parasitic Infections. Ed. Balows and Hausler. 6th edition, American Public Health Association, Washington, D.C.). Several reports indicate that C. jejuni enteritis is associated with eating in restaurants (Genigeorgis, C. 1987. pp. 111-145. In Elimination of Pathogenic Organisms from Meat and Poultry. Ed. F. J. M. Smulders, Elsevier, Amsterdam), drinking raw milk or unchlorinated water (Hutchinson et al. J. Hyg. Cam., vol. 94, pp. 204-215; Schmid et al. 1987. J. Infect. Dis., vol, 156, pp. 218-222), eating under-cooked poultry meat (Harris et al. 1986. Am. J. Pub. Health, vol. 76, pp. 407-411; Izat and Gardner. 1986. Poultry Science, vol. 67, pp. 1431-1435), and living in a household with pets (Genigeorgis, supra; Vandenberghe et al. 1982. Br. Vet. J., vol. 138, pp. 356-361). C. jejuni is common in the intestine of most domestic and many wild animals and is present in high numbers in most birds (Bokkenheuser, supra).
Species of Camplyobacter enter into a non-pathological, commensal relationship in the intestine of the chicken (Juven et al. 1991. Eur. J. Clin. Microbiol., v. 70, pp. 95-103). Chickens carry the organism as part of the indigenous intestinal flora, and for this reason they have been suspected as an important vehicle in the transmission or Campylobacter spp. to humans (Izat and Gardner, supra; Juven et al., supra). Poultry has been implicated as the major reservoir of human campylobacteriosis in the developed world (Genigeorgis, supra). As many as 90% of broiler chickens may harbor this organism (Lam et al. 1992. Av. Dis., v. 36, pp. 359-363). The pathogen can survive the chicken processing procedures and may be present on the product in retail outlets (Hood et al. 1988. Epidem. Inf., vol. 100, pp. 17-25). In a study at a live poultry market in New York City, more than 80% of broiler chickens sold yielded the organism. In another study, 68% of the retail broiler carcasses tested had C. jejuni (Izat et al., supra). Undercooked or improperly handled poultry has been implicated in 50 to 70% of Campylobacter spp. infections in humans (Hood et al., supra). In a study on university students in Georgia infected with Campylobacter, 70% of the cases were accounted for by eating chicken, often undercooked or raw, and 30% by contact with cats Tauxe et al. 1988. CDC Surveillance Summaries, vol. 37(SS-2), pp. 1-13!.
In a five year study conducted by the CDC between 1982 and 1986, 41,343 isolates of Campylobacter were reported. This yielded an annual isolation rate of 5.5 per 100,000 persons (Tauxe et al., supra). C. jejuni represented 99% of the reported species.
Campylobacter is believed to establish itself in the host by reaching and colonizing the mucosal surface of the intestinal tract. Enteropathogenic bacteria must overcome a number of host defense mechanisms to establish infection or to colonize. As part of the host defense mechanism, the mucous layer quickly eliminates unattached organisms from the gastrointestinal tract. As a virulence factor, the motility of the bacterium allows it to traverse the mucous layer and attach to or invade the epithelial cells of the intestinal tract (Newell and McBride. 1985. J. Hyg. Camb., vol. 95, pp. 217-227). Motility and chemotaxicity are considered important factors in the mechanism of association of C. jejuni with the intestinal epithelium (Griffiths and Park. 1990. J. Appl. Bacteriol., vol. 69, pp. 281-301). Field et al. (1981. Infect. Immun., vol. 33, pp. 884-892) showed, by scanning electron microscopy, that C. jejuni were present on, in, and below the mucous gel in the lower ileum of infected neonatal mice 2 h after inoculation. The ability to swim through environments of high viscosity, such as mucous, depends on the possession of a specialized type of motility.
To establish the role of flagella in the virulence of C. jejuni, Newell and McBride (supra) used a wild type strain and two non-motile variants (one flagellate and one non-flagellate) to conduct some colonization experiments. Their results showed that the aflagellate variant colonized the intestinal tract poorly. This occurred because the organisms were rapidly eliminated from the gut. The nonmotile flagellate colonized the gut as successfully as the wild type strain in some cases. These results suggest that flagella, active or inactive, are necessary for the efficient colonization of the gastrointestinal tract.
In a similar study (Morooka et al. 1985. J. Gen. Microbiol., vol. 131, pp. 1973-1980), the colonization of the intestinal tract by several mutant strains differed strikingly according to their motility. The wild type strain colonized well, while the aflagellated mutants where cleared. A poorly motile mutant, which had short flagellar filaments, colonized mice better than the non-motile flagellated mutants. These observations confirmed Newell's conclusion, but further suggested that motility was a necessary factor for the intestinal colonization by this pathogen.
Bacterial colonization of mucosal surfaces depends on the bacteria being able to maintain close proximity to the mucosa and to attach so as to avoid being swept away (Griffiths and Park, supra). C. jejuni colonizes the small intestine, mainly the ileum (Griffiths and Park, supra), but it may also colonize the colon. Invasiveness, enterotoxin, and cytotoxin production have all been implicated in causing campylobacteriosis (Genigeorgis, supra).
During infection, antibodies are made to a variety of Campylobacter surface structures, e.g. outer membrane proteins, and lipopolysaccharides (McSweegan et al. 1987. Infect. Immun., vol. 55, pp. 1431-1435). Nachamkin and Hart (1985. J. Clin Microbiol, vol. 21, pp. 33-38) did some Western blot analysis of the human antibody response to C. jejuni cellular antigens during gastrointestinal infections. They used acute and convalescent phase sera from patients, and they analyzed the antibody activity against their homologous infecting strains and heterologous clinical isolates. Their results showed that with acute phase sera, 3 major bands were recognized, one of which corresponded to the flagellar antigen. Convalescent phase sera recognized many more proteins and the Campylobacter flagellin was the major immunodominant component in all sera tested. The flagellin was not the major protein however in Coomassie blue stained gels.
Winsor et al. (1985. Gastroenterology, vol. 90, pp. 1217-1222) carried on some experiments to determine which C. jejuni outer membrane antigens elicited secretory IgA (sIgA) by using Western blot analyses of fecal extracts in patients with naturally acquired campylobacteriosis. Seven out of the eight patients elicited sIgA titres. The antigen to which the immunoglobulin reacted very strongly was the 63 kd flagellar antigen.
The flagellum is a major antigen of the Campylobacter cell (Harris et al. 1987. Am. J. Pub. Health, vol. 76, pp. 407-411), and it is the immunodominant antigen recognized during an infection in humans (Pavloskis et al. 1991. Infect. Immun., vol. 59, pp. 1159-1164). It has been reported that there were various classes of antibodies against the flagellar protein in convalescent sera (Ueki et al. 1988. Microbiol. Immunol., vol. 32, pp. 327-337). Herbrink et al. (1988. Eur. J. Clin. Microbiol. Infect. Dis., vol. 7, pp. 388-393) investigated the IgG, IgA, and IgM immune response against C. jejuni at various timepoints during and after infection in humans. Their results showed that IgG antibody titers generally remained at a constant level for more than 50 days, where IgA and IgM titers declined more rapidly to normal values within 30 to 50 days after onset of clinical symptoms.
When an isogenic aflagellar mutant was used to challenge a rabbit, the campylobacters were cleared in less than 24 h. There was no significant IgA response, and the non-flagellar mutant did not protect the rabbit against challenge with the parent strain (Pavloskis et al., supra).
Flagellar filament seems to carry some of the serogroup-specific epitopes, since non-flagellated mutants lose their capacity to be serotyped by the Lior procedure. For most LIO serogroups however, the contribution of the flagellum to serotypic specificity has yet to be determined (Harris et al., J. Bacteriol., supra). Flagella are the locomotory organelles of bacteria (Power et al. 1992. J. Bacteriol., vol. 174, pp. 3874-3883). They are reversible rotary devices, driven by protonmotive force that propel the bacteria through liquid environments (Macnab et al. 1991. Trends in Genetics, vol. 7, pp. 196-200). At a gross level, the known features of the flagellar apparatus are a filament, a hook, and a basal body. This structure is called the "filament hook basal-body complex" (Macnab and De Rosier. 1988. Can. J. Microbiol., vol. 34, pp. 442-451). The locations of the flagellar components fall into five compartments: the cytoplasmic face of the cell membrane, the cell membrane itself, the periplasmic space, the outer membrane, and the cell exterior. The hook is attached to the basal body. The hook and filament are both external to the cell. The flagellar filament is the portion that performs the hydrodynamic work on the cell's environment.
A flagellar filament is a long helical thread of uniform thickness. Its thickness is around 20 nm and its length is 15 .mu.m. Heating of flagellar filaments at 56.degree. C. for 15 min disintegrated them and released a single protein called flagellin (Iino, T. 1985. In Molecular Cytology of Escherichia coli, Academic Press, London, pp. 9-37). The MW of the flagellin monomer differs among different bacterial species, ranging from 40,000 to 63,000. C. jejuni flagellin monomer has a MW of 63,000 (Ueki et al. 1987. Microb. Immunol., vol. 31, pp. 1161-1171). The flagellin monomers, which formed globular units, are lined in 11 longitudinal rows, alternate with each other in adjacent rows, and form as a whole a tubular structure. Flagellin monomers at high concentration assemble by themselves and form filaments in vitro. The reaction is reversible, and the binding among the monomers is thought to be hydrophobic (Iino, supra).
The flagellar systems of similar bacteria, i.e. Escherichia coli (E. coli) and Salmonella typhimurium (S. typhimurium), are encoded by at least 40 genes organized into three regions on the chromosome (Mcnab, 1991, supra; Muller et al. 1992. J. Bacteriol., vol. 174, pp. 2298-2304). However, more than 60 genes are known to be involved in motility and chemotaxis (Macnab, R. M. 1987. In Escherichia coli and Salmonella typhimurium Cellular and Molecular Biology, Vol. 1., Eds. Neidhardt et al. American Society for Microbiology, Washington, D.C., pp. 732-759). The genes associated with the motile behavior are divided into three groups (Macnab, 1991, supra). Genes whose products are essential for the assembly of the flagella are given the symbol `fla`. Genes whose products are not necessary for the flagellar assembly, but essential for motor rotation, are given the symbol `mot`. The group of genes whose products are responsible for chemotactic responsiveness and control of switching between clockwise and anticlockwise direction of rotation, are given the symbol `che`. Nearly all of the flagellar, motility and chemotaxis genes are located in four clusters on the E. coli. chromosome (Macnab, 1991, supra).
The genes are organized into a number of operons, so regulation is especially critical with regards to the flagellin structural gene (Macnab and Aizawa. 1984. Ann. Rev. Biophys. Bioeng., vol. 13, pp. 51-83). A flagellar filament of typical length contains about 20,000 subunits. Synthesis of proteins in such large quantities is very wasteful if the bacterium cannot incorporate the proteins into the flagellum, due to a basal body defect for example. The majority of the regulatory mechanisms operate at the transcriptional level (Macnab, 1991, supra). They regulate expression of the flagellar genes in a hierarchy that parallels their roles in the assembly pathway. Operons coding for proteins needed in the initial steps of the assembly, i.e. switch, basal body, and export apparatus components, are expressed early. Genes for filament structure, motor rotation and chemotactic signaling, whose products are needed only when the basal body-hook complex is complete, are expressed late. All of the early genes must be expressed to obtain transcription of the late genes. A functional defect in any of the early genes can prevent expression of the late genes (Macnab 1991, supra).
The alteration of the transcriptional specificity of the RNA polymerase by the synthesis of alternative sigma factors provides a powerful way of controlling gene expression (Helmann, J. D. 1991. Molec. Microbiol., vol. 5, pp. 2875-2882). The flagellin protein accounts for greater than 98% of the mass of the bacterial flagellum. C. jejuni, among other enteric bacteria, was found to have a sigma-28-like promoter element preceding the flagellin genes (Helmann, supra). Another alternative sigma factor (sigma-54) was found to control flagellin expression in some bacteria. Campylobacters have two flagellin genes, flaA and flaB. A sigma-54-like promoter element was found upstream of the flaB gene, although only the sigma-28-dependent flaA protein is required for motility (Galan et al. 1990. Gene, vol. 94, pp. 29-35). The sigma factor is part of the control mechanisms over flagellin expression, the other mechanisms are still unknown.
The flagellin antigen is highly immunogenic (Khawaja et al. 1992. Curr. Microbiol., vol. 24, pp. 213-221). The flaA flagellin protein has been divided into three distinct regions consisting of two common and one variable regions (Fisher and Nachamkin. 1991. Molec. Microbiol., vol. 5, pp. 1151-1158). The two common regions, C1 and C2, comprising the N-terminal 170 amino acids and C-terminal 100 amino acids, showed 94% and 96% identity to Campylobactyer coli (C. coli) common flagellin regions, respectively. The variable V1 region, comprising the middle of the protein, shows 61% identity to C. coli residues. Comparison of these regions with the sequence of other bacteria, E. coil and Salmonella, showed a similar pattern but with much less identity.
The amino acid sequence of the flagellin N-terminal region, mainly the first 20 residues, has been shown to be homologous in all C. jejuni strains tested to date (Fisher and Nachamkin, supra). This part of the flagellin is essential for filament assembly. During assembly of the flagellum, flagellin subunits are transported through the center of the filament and polymerize at its tip (Nuitjen et al. 1990. J. Biochem. Chem., vol. 265, pp. 17798-17804). Both termini of the flagellum are important to the extension of the filament, and the amino terminus is necessary for the transport. By deletion analysis (Logan et al. 1989. J. Bacteriol., vol. 171, pp. 3031-3038), it was shown that the smallest E. coli flagellin capable of forming flagellar filament required the N-terminal 193 residues and the C-terminal 117 residues. The exposed antigenic regions are less restricted and susceptible to mutations, some of which are advantageous to the organism (Khawaja et al., supra).
Two copies of the flagellin gene of C. jejuni have been identified which are 95% identical (King et al. 1991. Microb. Ecol. Health Dis., vol. 4, pp. 135-140). Flagellar expression is subject to both phase and antigenic variation in Campylobacter species (Logan et al., supra), probably as an adaptation to the environment and the immune response of the host (Nuitjen et al. 1991. Infect. Immun., vol. 59, pp. 1100-1105). Phase variation refers to the ability of some strains to exhibit a bidirectional transition between flagellated and nonflagellated states (Guerry et al. 1990. J. Bacteriol., vol. 172, pp. 1853-1860). Antigenic variation refers to the ability of some strains to synthesize alternate flagellin protein that are distinguishable antigenically and that have different molecular weights. The immunogenicity and antigenic diversity of campylobacter flagella makes them important antigens in serotyping schemes based on the heat-labile antigens like the Lior scheme (Logan et al. 1987. J. Bacteriol., vol. 169, pp. 5072-5077). In some of the LIO serotypes the use of nonflagellated organisms has shown that the flagella can carry the serotype specific determinant (Logan et al., 1987, supra).
The two flagellin genes of C. jejuni 81116 were identified, cloned, and sequenced (Nuitjen et al., 1990, supra). The two copies of the flagellin genes were called flaA and flaB. Both genes are 1,731 base pairs each, they occurred as tandem repeats, and were 95% identical. They have the same orientation, and they are separated by a 173-bp intergenic region. The calculated moledular weights of flagellin A and B were 59,538 and 59,909, respectively. The estimated weight from polyacrylamide gels is 62,000; this difference is probably due to post translational modifications.
Nuitjen et al. (1990, supra) used two specific oligonucleotide probes to discriminate between the mRNA of flagellin A and B. In motile bacteria only mRNA transcribed from flagellin A was detected as a monocistronic messenger of about 1800 nucleotides. By carrying out primer extension studies on the mRNA, they located the start of transcription 43 nucleotides upstream of the ATG start codon. C. coli (Guerry et al. 1991. J. Bacteriol., vol. 173, pp. 4757-4764) also have two copies of the fla gene, flaA and flaB. The two genes share 91.9% sequence identity. Both products are expressed and are required for a fully active flagella (Wassenaar et al. 1991. EMBO J., vol. 10, pp. 2055-2061).
Harris et al. (1987, supra) showed that the flagella of certain strains of C. jejuni and C. coli undergo antigenic variation. C. jejuni 81116 expressed one of two flagellin proteins, one with a MW of 61,500 and the other with a MW of 60,000. A reversible DNA rearrangement has been detected in a C. coli strain, but not in C. jejuni (Harris, 1987, supra). King et al. (supra) studied the expression of flagellin with isolates associated with a milk-borne outbreak of campylobacteriosis. They found that the milk isolates expressed a flagellin with a MW of 62,000 while the human isolates expressed a 58,000 flagellin. They speculated that this antigenic variation gave a virulence advantage for the phenotype.
Very few C. jejuni genes have been cloned and expressed in E. coli. This is due mainly to the lack of genetic markers, the absence of a developed natural gene transfer mechanism, and possibly due to some distinct differences in the regulatory sequences of these two bacteria (Chan et al. 1988. Gene, vol. 73, pp. 185-191). Two genes that have been expressed in E. coli are proB (gamma-glutamylkinase) and proA (gamma-glutamylphosphate-reductase). These genes were isolated by complementation of pro mutant E. coli. It is speculated that these genes were expressed only because the host cells were under pressure. Some of the genes identified in C. jejuni are glyA gene (serine hydroxymethyltransferase) (Chan and Bingham. 1990. Gene, vol. 101, pp. 51-58; Chan, 1988, supra), lysyl-tRNA sypthetase gene (Chan and Bingham. 1992. J. Bacteriol., vol. 174, pp. 695-701), and the 5S, 16S and 23S ribosomal RNA (Ouellette et al. 1987. Antimicrob. Agents Chemother., vol. 31, pp. 1021-1026).
Wang and Taylor (1990. J. Bacteriol., vol. 172, pp. 949-955) reported that growing cells of C. jejuni and C. coli could be naturally transformed by naked DNA without the requirement for any special treatment. Maximum competence was found in early log phase of growth. The cells took up their own DNA much better than E. coli DNA.
Recently Labine-Roussel et al. (1987. J. Bacteriol., vol. 169, pp. 5320-5323) constructed a shuttle cloning vector which can be mobilized from E. coli to C. jejuni, C. coli, and Campylobacter fetus (C. fetus). This vector was used to carry on gene disruption and replacement via homologous recombination (Labigne-Roussel et al. 1988. J. Bacteriol., vol. 170, pp. 1704-1708).
The host responses to intestinal microbial infections involves a complex interplay of soluble factors or mediators, leukocytes, epithelial and endothelial cells of the gut-associated lymphoid tissue (GALT). The GALT is one component of the mucosa-associated lymphoid tissue (MAST), which also includes the bronchial, salivary, nasopharyngeal and genitourinary lymphoid tissues. The GALT consists of discrete lymphoid follicles scattered along the wall of the small intestine (Mesteky and McGhee. 1987. Adv. Immunol., vol. 40, pp. 153-229).
The GALT in chickens consists of the bursa of Fabricius, cecal tonsils (CT), Peyer's patches (PP), and lymphocyte aggregates in the intraepithelium and in the lamina propria (LP) of the gastrointestinal wall. The bursa of Fabricius was considered to be the only site where antibody-forming cells could form (Befus et al. 1980. J. Immunol., vol. 125, pp. 2626-2632). However, surgical ablation of the bursa of Fabricius, even in early embryonic development, does not completely inhibit the production of a humoral response. Thus, other non-bursal lymphoid tissue support some B cell differentiation (Befus et al., supra).
It has been suggested that prevention of infection by C. jejuni can be attained by blocking the colonization factor with specific antibodies (Ueki et al., supra). Wu et al. (1991. Infect. Immun., vol. 59, pp. 2555-2559) showed that the flagellar protein was the major antigen recognized by intestinal lavage IgA in mice infected with C. jejuni.
Serum antibody response to invasive enteric pathogens is very important in protection against systemic infections. The initial immunologic response to enteric infection occurs at the level of the intestinal mucosa. Secretory immunoglobulin A (sIgA) response at the intestinal mucosa is a primary defense against enteric infections (Winsor et al. supra). Stern et al. (1990. Avian Dis., vol. 34, pp. 595-601) found that specific anti-C. jejuni antibodies diminish the ability of the bacterium to colonize the gut of 1-day-old chicks when incubated with the organism as compared with preincubation with phosphate buffered saline.
The flagella of C. jejuni are essential in the colonization of the intestine. Nonflagellated organisms are quickly cleared from the intestine. Chicken polyclonal antiflagellin antibodies as well as monoclonal antiflagellin antibodies have been found to prevent C. jejuni from colonizing the chickens or to increase the dose of bacteria required to colonize the chickens (Carr, unpublished). Flagellar antigens are therefore potential candidates for vaccines as well as suitable antigens for diagnostic purposes, since the flagellin protein is immunodominant during human infections.
Kim et al. (1989. Infect. Immun., vol. 57, pp. 2434-2440) immunized chickens with live E. coli expressing Eimeria acervulina merozoite recombinant antigen. The transformant cells were administered orally. Their results suggested that the recombinant vaccine could elicit antigen-specific humoral and cellular immune responses against the protozoan. Challenge with infective oocysts enhanced both immune responses, implying that the vaccine primed the chicken immune system against this protozoan. The protection, however, was partial. Immunoglobulin and T-cell responses against the recombinant antigen could be detected 7 days after vaccination.
Oral immunization to induce immunity against infectious diseases is convenient, relatively safe, and takes advantage of the mass of lymphoid tissue associated with the gut (Liang et al. 1988. J. Immun., vol. 141, pp. 1495-1501).
The protective role of sIgA is well documented in many experimental models. sIgA neutralizes viruses, toxins, enzymes, inhibits adherence of bacteria to epithelial surfaces (Mesteky and McGhee, supra). sIgA binds to and agglutinates bacteria, but it is not thought to be bateriocidal (McSweegan et al., supra). Thus the induction of specific sIgA is desireable to selectively inhibit and clear colonizing bacteria from the gut. The presence of antibody-antigen complexes in the gut is known to stimulate the production of large quantities of mucus. This flow of mucus will trap the pathogens which will be more readily removed by normal intestinal peristalsis. Moreover, sIgA are better adapted in secretions, being more resistant to denaturation and proteolytic breakdown than IgG antibodies (Pierre et al. 1988. Immunology, vol. 18, pp. 51-56).
The major natural pathway for stimulating the immune system is thus through the GALT, where natural or artificially introduced antigens penetrate through the highly pinocytic and phagocytic M cells and interact with resident accessory and lymphoid cells (Mesteky and McGhee, supra). Precursor IgA B cells leave the site, mature and home back to the lamina propria of the GALT where they differentiate into IgA plasma cells specific for ingested antigens (Mesteky and McGhee, supra).